A PRIORI DECIDED PROCEDURE for ‘ruff sperm’ project
Martin Bulla, Wolfgang Forstmeier, Katrin Martin, Kim Teltscher, ?Jasmine L Loveland?, Clemens Küpper, Michael Lierz, Tomas Albrecht, David B Lank & Bart Kempenaers
2021-03-08 09:38:18
Code to load tools and prepare data:
require(here)
source(here::here('R/tools.R'))
colors <- c("#999999", "#E69F00", "#56B4E9") #viridis(3)
v = data.table(read_excel(here::here('Data/ruff_sperm_Vancouver_2018.xlsx'), sheet = 1))
v = v[!is.na(sample_ID)]
x = v[!duplicated(bird_ID)]
s = data.table(read_excel(here::here('Data/ruff_males_Seewiesen.xlsx'), sheet = 1))
setnames(s, 'Morph', 'morph')
sv = x[bird_ID%in%s$Ind_ID]Background
Ruff Calidris pugnax is a lekking shorebird with three strikingly different mating morphs: aggressive ‘independents’, semi-cooperative ‘satellites’ and female-mimic ‘faeders’ (vanRhijn 1991; Widemo 1998; Jukema and Piersma 2006). ‘Independents’ (80 to 95% of the population) are dominant holders of display sites. Satellites (5 to 20%) do not defend display sites, but their presence on the lek assists with female attraction and allows them to ‘steal’ copulations. ‘Faeders’ (<1%) mimic females in their plumage and smaller size and also ‘steal’ copulations. The major differences in body and testis size, ornamentation, and mating behaviours of ‘satellites’ and ‘faeders’ develop via an autosomal inversion (Kupper et al. 2015; Lamichhaney et al. 2015). Whether the morphs also differ in sperm morphology and velocity has not been investigated.
Research questions
Predictions
- Do the three ruff morphs differ in the sperm morphology and velocity, as well as in the within- and between-individual variability of the sperm traits?
The morphs may not differ in sperm morphology and velocity because the morphs differ little in spermatogenesis-related gene expression in testis (Loveland et al. 2021).
set.seed(1)
d = data.table(morph = factor(c( rep('independent', 60),
rep('satelite', 30),
rep('faeder', 15)
),
levels = c('independent','satelite','faeder')
),
sperm_trait = c(rnorm(60, mean = 1, sd = 0.5),
rnorm(30, mean = 2, sd = 0.5),
rnorm(15, mean = 3, sd = 0.5)
)
)
#dev.new(width = 3.5, height = 3.5)
ggplot(d, aes(x = morph, y = sperm_trait, col = morph)) +
geom_jitter() + theme_MB + theme(legend.position="none") + xlab('Morph') + ylab('Sperm length or relative midpiece length') +
scale_colour_manual(values = colors)
On the other hand, the three morphs may differ in sperm traits because sperm length and swimming speed as well as investment in sperm numbers (testes size) is generally related to the intensity of sperm competition [CIT]. Indeed, ‘independents’ have relatively smaller testes then ‘faeders’ and ‘satelites’ (Kupper et al. 2015). In contrast, dominant ‘independents’ on the lek obtain the highest number of copulations, the likelihood of copulation for other ‘independents’ and ‘satellites’ is driven by the amount of time an individual spends on the lek (Vervoort and Kempenaers 2019). Importantly, investing in pre-copulatory male-male competition may trade-off with investing in post-copulatory male-male competition. We thus predict that ‘faeders’ have the longest sperm, midpiece and tail and have sperm that is fastest of the three morphs (Figure 1 & 2). As faders invest little in pre-copulatory male-male competintion, we also expect ‘faeders’ sperm to differ the most from the sperm of ‘independents’, while we expect ‘satelites’, who partly invest in pre-copulatory male-male competition, but to a lesser extant than ‘independents’ (Vervoort and Kempenaers 2019), to have intermediate sperm between the two (Figure 1). Consequently, we predict the smallest coefficient of variation in the sperm morphometry ?and velocity? in ‘faeders’ as within- and between-individual variability in sperm traits is lower in species with higher sperm competition (Kleven et al. 2008; Lifjeld et al. 2010). This expectation is based on the assumption that the sperm production is better optimized and less error prone in ‘faeders’ than in the other two morphs. Such optimizations shall increase ‘faeders’ chances of inseminating a female despite ‘faeders’ only sporadic chances to copulate [CIT NEEDED or DELETE].
- Does sperm morphology predict sperm swimming speed? Sperm morphology, specifically long sperm, sperm with short heads, and long midpiece (relative to overall sperm length), has been positively associated with sperm swimming speed. We thus predict that longer sperm and/or sperm with longer head, and/or sperm with relatively longer midpiece will swim faster.
Figure 1 | Predicted relationship between morph type and sperm trait.
set.seed(1)
d = data.table(morph = c( rep('independent', 60),
rep('satelite', 30),
rep('faeder', 15)
),
midpiece = c(rnorm(60, mean = 2, sd = 0.5),
rnorm(30, mean = 3, sd = 0.5),
rnorm(15, mean = 4, sd = 0.5)
),
tail = c(rnorm(60, mean = 2, sd = 0.5),
rnorm(30, mean = 3, sd = 0.5),
rnorm(15, mean = 4, sd = 0.5)
),
velocity = c(rnorm(60, mean = 2, sd = 0.5),
rnorm(30, mean = 3, sd = 0.5),
rnorm(15, mean = 4, sd = 0.5)
)
)
d[,morph123 :=ifelse(morph == 'independent', 1, ifelse(morph == 'satelite', 2,3))]
colors_ <- colors[d$morph123]
par(las = 1, cex.axis = 0.6, cex.lab = 0.8, cex.main = 0.8)
s3d=scatterplot3d(d$midpiece, d$tail, d$velocity, pch = 16, type="h",
color=colors_, grid=TRUE, box=FALSE,
xlab = "",
ylab = "",
zlab = "",
x.ticklabs=c("short","","","","","long"),
y.ticklabs=c("short","","","","long",""),
z.ticklabs=c("slow","","","","","fast"),
mar = c(3, 2, 0, 1.5)
)
text(x = 7.5, y = 1, "Tail", srt = 0, cex = 0.8)
text(x = 2.5, y = -0.5, "Mipiece", srt = 0,xpd = TRUE, cex = 0.8)
text(x = -0.5, y = 2.5, "Velocity", srt = 90,xpd = TRUE, cex = 0.8)
legend("bottom", legend = levels(factor(d$morph,levels = c('independent','satelite','faeder'))),
col = colors, pch = 16,xpd = TRUE, horiz = TRUE,inset = -0.125, bty = "n", cex = 0.7)
Figure 2 | Predicted relationship between morph type and sperm traits.
Methods
Individuals
In 2018, we collected and stored in ~5% formalin the vas deferens from 15 sacrificed males (5 per morph) from the Simon Fraser University colony. The colony was founded with 110 ruffs hatched from wild eggs collected in Finland in 1985, 1989 and 1990 plus two ‘faeder’ males from the Netherlands in 2006 (Lank et al. 1995, 2013).
In 2021, we plan to collect sperm of all males from the Max Planck colony (64 ‘independents’, 28 ‘satelites’, 8 ‘faeders’; of which 40 males were sampled already in Vancouver: 26 ‘independents’, 10 ‘satelites’ and 4 ‘faeders’). The colony was founded in 2018 from the Simon Fraser colony.
Sperm sampling
Compared to passerines, shorebirds do not have a cloaca protuberance. Thus, abdominal massage and cloacal lavage method of sperm collection (e.g. Knief et al. (2017) ; see the specific protocol) often lead to unclean samples contaminated by excrements. Such samples are then unsuitable for measuring sperm velocity. Hence, in 2021 along with abdominal massage and cloacal lavage, we plan to collect sperm by electro‐stimulation (Lierz et al. 2013). The length and diameter of the electro-stimulation probe, as well as the electric current and the number of electric impulses will be adapted to the sampled individuals. The electro-stimulation should result in clean sperm samples suitable for velocity measurements.
The morphs will be sampled in a haphazard order to avoid temporal trends in morph sampling. In other words, all males of one morph will not be sampled at the same time. Sperm will be pipetted from the cloaca (~0.5–3μl) and immediately diluted in a preheated (40°C) Dulbecco’s Modified Eagle’s Medium (Advanced D-MEM, Invitrogen, USA). For the velocity measurements, an aliquot will be pipetted onto a standard 20μm two-chamber count slide (Leja, The Netherlands) placed on a thermal plate (Tokai Hit, Tokai Hit Co., LtD., Japan) kept at 40 °C. For the morphology measurements, an aliquot will be pipetted onto a microscopy slide and the rest of the sperm sample will be fixed in 100μl ~5% formalin solution.
Sperm morphometry
From each sperm sample, we will pipet ~10μl onto a microscopy slide, let it dry at room temperature, and - in case of formalin-fixed samples - wash the slide gently under distilled water to wash away the dried formalin and phosphate-buffered saline solution. Sperm from the vas deferens will be first extracted by cutting a piece from the middle of the tubules, plucking apart the tissue with curved extraction forceps to release the sperm, and washing away the sperm with the phosphate phosphate-buffered saline solution.
We will inspect the slides with a light microscope (‘Zeiss Axio Imager.M2’) under 200x magnification. To aid distinction of sperm parts, dried slides will be stained with Hoechst 33342 (which identifies the sperm nucleus) and Mitotracker Green FM (which identifies sperm midpiece) not earlier than 48h before photographing. For step by step slide preparation protocol see here.
Picture 1 | Ruff sperm blue-stained nucleus by Hoechst 33342 and green-stained midpiece by Mitotracker Green FM
For each sample, we will photograph at least 10 intact normal-looking spermatozoa, with a 12 megapixel (4250 × 2838) digital camera (‘Zeiss Axiocam 512 color’ with pixel size of 3.1μm × 3.1μm). Ten sperm per male were previously used to investigate the relationship between coefficient of variation in sperm trait and sperm competition (Kleven et al. 2008; Lifjeld et al. 2010). For each sperm, we will measure the length of the acrosome, the nucleus, the midpiece and the tail to the nearest 0.1μm using the open source software ImageJ (for a detailed sperm measuring protocol see here). To minimize observer error, all measurements will be taken by one person (KT), who will be blind to the individual morph and will measure the sperm from all morphs and samples in a random order (NOTE THAT THIS IS NOT THE CASE FOR THE VAS DEFERENS).
We will calculate total sperm length as the sum of all parts, head length as the sum of acrosome and nucleus length, and flagellum length as the sum of midpiece and tail length. We will compute coefficients of variation for each trait (CV = [SD/mean]) both within males and between males of each morph. If we are unable to measure 10 sperm for all males, we will correct the coefficients of variation for variation in sample size (CVadj = [1 + 1/(4n)]*CV; Sokal and Rohlf (1981)).
Each measured sperm and sperm part will be numbered and referenced in the database (see sperm measuring protocol and database). This ensures transparency and allows re-measurement of the same sperm by the same or different person.
Sperm velocity
For each sperm sample we will record sperm velocity for approximately 45s in eight different fields of the slide under a 100x magnification using a phase contrast and a digital camera (UI-1540-C, Olympus) mounted on a microscope (CX41, Olympus) fitted with a thermal plate (Tokai Hit, Tokai Hit Co., LtD., Japan) kept at a constant temperature of 40°C. The videos will be recorded at 25 frames per second. A single person will analyze each recorded field using the CEROS computer-assisted sperm analysis system (Hamilton Thorne Inc., Beverly, Massa- chusetts, USA), visually inspect the tracked objects and exclude non-sperm objects and static spermatozoa from the analysis (Laskemoen et al. 2010; Cramer et al. 2016; Opatová et al. 2016). The dilution medium does not contain any spermatozoa attractants to guide the spermatozoa towards one direction; thus, we will use curvilinear velocity rather than straight-line velocity as our measurement of sperm swimming speed (Laskemoen et al. 2010). We will report how many sperm cells per sperm sample will be recorded (median, 95%CI, range) and - if necessary - control for the (log-transformed) number of measured sperm cells per sample.
Inbreeding
Because our sperm samples came and will come from males bred in captivity, we expect higher levels of inbreeding compared with males from wild populations. In birds and mammals (but not in insects) inbred males have a higher proportion of abnormal sperm and lower sperm velocity than outbred males (Gomendio, Cassinello, and Roldan 2000; Heber et al. 2013; Ala-Honkola et al. 2013; Opatová et al. 2016). Thus, our sperm velocity measurements may not reflect sperm velocity of wild ruffs. However, there is no evidence that the morphology of normal‐looking sperm (e.g., length, coefficient of variation) differs between inbred and outbred males (Mehlis et al. 2012; Ala-Honkola et al. 2013; Opatová et al. 2016). Based on these studies, we assume that our measurements reflect the variation in sperm morphology observed in wild ruffs. Nevertheless, we know the relatedness of individuals within our population (pedigree) and can estimate inbreeding from microsatellite markers. Hence, we will investigate whether variation in inbreeding and genetic diversity is linked to the sperm traits and if so we will attempt to control for it in our models. Similarly, if find between morph differences in sperm traits, we will investigate whether such differences might have occurred by chance due to population bottlenecks.
Analysis plan
Sample sizes will reflect the maximum available data. No data selection will be done conditional on the outcome of statistical tests. To avoid unconscious data dragging, we will first analyze the data on sperm traits blind to the males morphs. Specifically, we will first randomly assign male morph to the measurements, explore such data and finalize the analyze using such data. Only then we will run the analyses with true assignment of the morphs. We will report all results, all data exclusions, all manipulations and all measures in the study at the GitHub repository.
Statistical analyses
All analyses will be in R (R-Core-Team 2020) using lmer() and glmer() functions of the lme4 package to fit linear and generalized linear mixed-effects models, and - if necessary - one of the package for fitting the pedigree structure as a random effect (e.g. brms, MCMCglmm, pedigreeMM). We will estimate the variance explained by the fixed effects of our mixed-effects models as marginal R2-values, using the r.squaredGLMM() function of the MuMIn package (v1.15.6). Model fits will be visually validated (by qq-plots of residuals and plots of residuals against fitted values).
In general, we will fit two sets of linear models. A first set of models will have as a response variable the raw sperm trait measurements and will be controlled for multiple measurements per male by fitting male identity as a random effect. A second set of models will have average sperm trait per male as a response and - if necessary - will be controlled for the number of sampled sperm per individual by specifying a wights term as sqrt(n-3)(cit needed). MB THE FIRST SET OF MODLES IS RATHER COMPLEX (if pedigree is used), THAT IS WHY I PROPOSE ALSO THE SIMPLER AVERAGE VARIANT. THE AVERAGE VARIANT WILL ALSO CORRESPOND WITH MODELS FOR VELOCITY. HOWEVER, IF THIS MAKES THINGS TO COMPLICATED, WE CAN RUN JUST THE FIRST SET.
Yes, weights (e.g. sqrt(n-3)) would be possible, but I would really to try keep n constant, and maybe rather exclude the 1-3 individuals for which we cannot obtain 10 sperm, since such ‘ejaculates’ may be weird anyway.
I agree with Bart in principle, but I presume that such an analysis will be more illustrative about the amount of variance explained by morph, and also better comparable to analyses of CV.
To investigate whether sperm traits differ between the morphs, we will fit a model for each sperm trait separately, with the male morph (three-level factor) as an explanatory variable, controlling for the aviary, in which an individual is housed (aviary ID), and - if necessary - controlling for the pedigree structure.
To investigate whether sperm morphology explains variation in sperm velocity, we will first fit a linear model with average velocity per male as the dependent variable, and with male-average total sperm length, head (acrosome + nucleus), midpiece and tail length (?or rather flagelum - midpiece+tail?), as well as their squared terms (after mean-centering) and all two-way interactions between the three linear terms, as explanatory variables. If necessary, we will control for a pedigree structure. In case, some of the sperm traits substantially correlate (Pearson’s r>0.6), we will use only one of the correlated traits. We will then investigate whether predicted sperm velocity corresponds with observed sperm velocities.
THE FOLLOWING IS LIKELY NOT NEEDED: To evaluate multicollinearity between all main effect predictors, we will estimate their variance inflation factor using the corvif() function (Zuur, Ieno, and Elphick 2009) in R. A general guideline is that a variance inflation factor larger than 5 or 10 is large, indicating that the model has problems estimating the coefficient. However, this in general does not degrade the quality of predictions. If the VIF is larger than 1/(1-R2), where R2 is the Multiple R-squared of the regression, then that predictor is more related to the other predictors than it is to the response. Thus, if VIF is larger than 5 we can use the model for predictions.
Proposed time line
- ongoing - measuring sperm morphology from vas deferens samples
- May - data collection
- June - July - photographing and measuring sperm samples, quantifying sperm velocity
- August - September - analyses and 1st draft
TO DECIDE
- Shall we add the good 2018 samples, especially from faeders?
- Is it possible to add some of Clemens birds, e.g. faeders?
- Does it matter whether we let the microscope slides dry at room temperature or whether we heat fix them?
- Shall we use average sperm measurements in all analyses, given that for velocity-morphology relationship we will use average values?
- What sperm traits to use? Total length with/without acrosome, acrosome, head (acrosome + nucleus), midpiece, tail, flagelum (midpiece + tail). BART: total length, head, midpiece and tail
- Shall we investigate whether the sperm measurements are representative of each male and morph by calculating the repeatability of sperm measurements per male and morph, which will be obtained through 1,000 parametric bootstrap iterations (Stoffel et al. 2017). BART: Yes.
- Shall we use discriminant function analyses on the two most different sperm traits to predict the sperm morph? BART: Perhaps use all sperm traits, calculate PCA and plot PC1 vs PC2 with morph in different colours? Not sure this is necessary, but might be a nice way to illustrate the differences (if any)?
- Do we need to control for the pedigree, if residuals of the models are randomly distributed, i.e. there is no pedigree structure in residuals?
- Which packages to use for modeling: brms, MCMCglmm, pedigreeMM
- What is the sound way to control for number of measured sperm per individual (in case this is variable) - weights? (Wolfgang?)
- When recording sperm velocity, shall we also make an additional record in a cellulose solution - a viscous environment, which “mimics” conditions within the female reproductive tract (Schmoll et al. 2020)? BART: No, if it compromises number of sampled individuals. a viscous environment
- Shall we look at occurrence of aberrant/abnormal sperm?
TO ADD/SPECIFY
- section on Housing of the birds
- specification of the heating table and video resolution (Tomas)
- videos to accompany the protocols (vas deferens preparation, slide preparation, sperm collection and video recording)
References
Ala-Honkola, O., D. J. Hosken, M. K. Manier, S. Lupold, E. M. Droge-Young, K. S. Berben, W. F. Collins, J. M. Belote, and S. Pitnick. 2013. “Inbreeding Reveals Mode of Past Selection on Male Reproductive Characters in Drosophila Melanogaster.” Ecol Evol 3 (7): 2089–2102. https://doi.org/10.1002/ece3.625.
Cramer, E. R., M. Alund, S. E. McFarlane, A. Johnsen, and A. Qvarnstrom. 2016. “Females Discriminate Against Heterospecific Sperm in a Natural Hybrid Zone.” Evolution. https://doi.org/10.1111/evo.12986.
Gomendio, M., J. Cassinello, and E. R. Roldan. 2000. “A Comparative Study of Ejaculate Traits in Three Endangered Ungulates with Different Levels of Inbreeding: Fluctuating Asymmetry as an Indicator of Reproductive and Genetic Stress.” Proc Biol Sci 267 (1446): 875–82. https://doi.org/10.1098/rspb.2000.1084.
Heber, S., A. Varsani, S. Kuhn, A. Girg, B. Kempenaers, and J. Briskie. 2013. “The Genetic Rescue of Two Bottlenecked South Island Robin Populations Using Translocations of Inbred Donors.” Proc Biol Sci 280 (1752): 20122228. https://doi.org/10.1098/rspb.2012.2228.
Jukema, Joop, and Theunis Piersma. 2006. “Permanent Female Mimics in a Lekking Shorebird.” Biology Letters 2 (2): 161–64. https://doi.org/10.1098/rsbl.2005.0416.
Kleven, Oddmund, Terje Laskemoen, Frode Fossøy, Raleigh J. Robertson, and Jan T. Lifjeld. 2008. “Intraspecific Variation in Sperm Length Is Negatively Related to Sperm Competition in Passerine Birds.” Evolution 62 (2): 494–99.
Knief, Ulrich, Wolfgang Forstmeier, Yifan Pei, Malika Ihle, Daiping Wang, Katrin Martin, Pavlína Opatová, et al. 2017. “A Sex-Chromosome Inversion Causes Strong Overdominance for Sperm Traits That Affect Siring Success.” Nature Ecology & Evolution 1 (8): 1177–84. https://doi.org/10.1038/s41559-017-0236-1.
Kupper, C., M. Stocks, J. E. Risse, N. Dos Remedios, L. L. Farrell, S. B. McRae, T. C. Morgan, et al. 2015. “A Supergene Determines Highly Divergent Male Reproductive Morphs in the Ruff.” Nat Genet. https://doi.org/10.1038/ng.3443.
Lamichhaney, S., G. Fan, F. Widemo, U. Gunnarsson, D. S. Thalmann, M. P. Hoeppner, S. Kerje, et al. 2015. “Structural Genomic Changes Underlie Alternative Reproductive Strategies in the Ruff (Philomachus Pugnax).” Nat Genet. https://doi.org/10.1038/ng.3430.
Lank, David B., Lindsay L. Farrell, Terry Burke, Theunis Piersma, and Susan B. McRae. 2013. “A Dominant Allele Controls Development into Female Mimic Male and Diminutive Female Ruffs.” Biology Letters 9 (6). https://doi.org/10.1098/rsbl.2013.0653.
Lank, David B., Constance M. Smith, Olivier Hanotte, Terry Burke, and Fred Cooke. 1995. “Genetic Polymorphism for Alternative Mating Behaviour in Lekking Male Ruff Philomachus Pugnax.” Nature 378 (6552): 59–62. https://doi.org/10.1038/378059a0.
Laskemoen, T., O. Kleven, F. FossÃy, R. J. Robertson, G. Rudolfsen, and J. T. Lifjeld. 2010. “Sperm Quantity and Quality Effects on Fertilization Success in a Highly Promiscuous Passerine, the Tree Swallow Tachycineta Bicolor.” Behavioral Ecology and Sociobiology, 1–11. http://www.scopus.com/inward/record.url?eid=2-s2.0-77951236330&partnerID=40&md5=4cbc6dbbaec0e670e386b9274a5dcf70.
Lierz, M., M. Reinschmidt, H. Muller, M. Wink, and D. Neumann. 2013. “A Novel Method for Semen Collection and Artificial Insemination in Large Parrots (Psittaciformes).” Sci Rep 3: 2066. https://doi.org/10.1038/srep02066.
Lifjeld, Jan T., Terje Laskemoen, Oddmund Kleven, Tomas Albrecht, and Raleigh J. Robertson. 2010. “Sperm Length Variation as a Predictor of Extrapair Paternity in Passerine Birds.” PLoS ONE 5 (10): e13456. https://doi.org/10.1371%2Fjournal.pone.0013456.
Loveland, J. L., L. M. Giraldo-Deck, D. Lank, W. Goymann, M. Gahr, and C. Küpper. 2021. “Functional Differences in the Hypothalamic-Pituitary-Gonadal Axis Are Associated with Alternative Reproductive Tactics Based on an Inversion Polymorphism.” Hormones and Behavior 127. https://doi.org/https://doi.org/10.1016/j.yhbeh.2020.104877.
Mehlis, Marion, Joachim G. Frommen, Anna K. Rahn, and Theo C. M. Bakker. 2012. “Inbreeding in Three-Spined Sticklebacks (Gasterosteus Aculeatus L.): Effects on Testis and Sperm Traits.” Biological Journal of the Linnean Society 107: 510–20. https://doi.org/10.1111/j.1095-8312.2012.01950.x.
Opatová, P., M. Ihle, J. Albrechtová, O. Tomášek, B. Kempenaers, W. Forstmeier, and T. Albrecht. 2016. “Inbreeding Depression of Sperm Traits in the Zebra Finch Taeniopygia Guttata.” Ecol Evol 6 (1): 295–304. https://doi.org/10.1002/ece3.1868.
R-Core-Team. 2020. “R: A Language and Environment for Statistical Computing - Version 4.0.2.” Vienna (Austria): R Foundation for Statistical Computing. http://www.R-project.org/.
Schmoll, T., G. Rudolfsen, H. Schielzeth, and O. Kleven. 2020. “Sperm Velocity in a Promiscuous Bird Across Experimental Media of Different Viscosities.” Proc Biol Sci 287 (1931): 20201031. https://doi.org/10.1098/rspb.2020.1031.
Sokal, R., and F. Rohlf. 1981. Biometry: The Principles and Practice of Statisticsin Biological Research (2nd Ed.). New York, NY: W. H. Freeman and Co.
Stoffel, Martin A., Shinichi Nakagawa, Holger Schielzeth, and Sarah Goslee. 2017. “rptR: Repeatability Estimation and Variance Decomposition by Generalized Linear Mixed‐effects Models.” Methods in Ecology and Evolution 8 (11): 1639–44. https://doi.org/10.1111/2041-210x.12797.
vanRhijn, Johan G. 1991. The Ruff. T. & A.D. Poyser, London.
Vervoort, Raf, and Bart Kempenaers. 2019. “Variation in Lek Attendance and Copulation Success of Independent and Satellite Male Ruffs Calidris Pugnax.” Ardea 107. https://doi.org/10.5253/arde.v107i3.a9.
Widemo, F. 1998. “Alternative Reproductive Strategies in the Ruff, Philomachus Pugnax: A Mixed ESS?” Animal Behavior 56: 329–36. https://doi.org/https://doi.org/10.1006/anbe.1998.0792.
Zuur, Alain F., Elena N. Ieno, and Chris S. Elphick. 2009. “A Protocol for Data Exploration to Avoid Common Statistical Problems.” Methods in Ecology and Evolution 1 (1): 3–14. https://doi.org/10.1111/j.2041-210X.2009.00001.x.